大跨径钢桥沥青混合料铺装路面缺陷检测算法改进  

Improvement of pavement defect detection algorithm for asphalt mixture pavement of long-span steel bridge

作　　者：刘吉昌 LIU Jichang(China Railway 14th Bureau Group First Engineering Development Co.,Ltd.,Rizhao 276800,Shandong China)

机构地区：[1]中铁十四局集团第一工程发展有限公司,山东日照276800

出　　处：《粘接》2025年第2期95-98,共4页Adhesion

摘　　要：以往的路面缺陷检测算法仅提取了路面图像中的边缘特征,导致算法的检测效果不佳。为此,本研究针对大跨径钢桥面沥青混合料铺装路面,设计了新的缺陷检测算法。在利用相机采集大跨径钢桥面沥青混合料铺装路面图像后,对其进行去噪和均衡化处理。然后,在对路面图像数据进行增强处理的基础上,利用边缘检测提取出路面图像的边缘特征,通过纹理分析提取纹理特征,并对2者进行融合处理。最后,通过计算特征阈值,对提取的特征进行分类,筛选出其中的路面缺陷特征,由此生成对应的路面缺陷检测算法。经实验测试验证,该算法在实际应用中的漏检率较低,检测效果较好。The previous pavement defect detection algorithm only extracted the edge features in the pavement image,resulting in poor detection effect of the algorithm.Therefore,a new defect detection algorithm was designed for the asphalt mixture pavement of long-span steel bridge deck.After the camera was used to collect the image of the asphalt mixture pavement of the large-span steel bridge deck,the image of the asphalt mixture pavement was denoised and equalized.Then,on the basis of the enhanced processing of the pavement image data,the edge features of the pavement image were extracted by edge detection,and the texture features were extracted through texture analysis,and the two were fused.Finally,by calculating the feature threshold,the extracted features were classified,and the pavement defect features were screened out,and the corresponding pavement defect detection algorithm was generated.Experimental tests show that the algorithm has a low missed detection rate and a good detection effect in practical applications.

关 键 词：大跨径钢桥面 沥青混合料 桥面铺装 缺陷检测 算法设计 

分 类 号：TQ321.3[化学工程—合成树脂塑料工业] U416.217[交通运输工程—道路与铁道工程]